Summicron-R versus Summicron-M

[imxputs@knoware.nl (Erwin Puts/imX)]

A number of readers of the magazine I am writing for asked me to
investigate the differences (if any) between the two standard 50mm lenses
for the R and M.As you may know, both lenses use the same optical formula,
but the R lens is physically larger, the lensdiameter is bigger an it is
slightly retrofocus, in order to accomodate the 47mm distance from
lensflange to film (the M has 27,8mm).  I started this research with a full
two week tour (on holiday) shooting all scenes (day and night, buildings,
animals, landscapes, persons, closeups and far distance shots, in the sun
and on cloudy days), twice (with both lenses that is). I made sure my
shutterspeed was never below 1/125 or using a tripod. I ended up with more
than 400 transparancies on Agfa RSX50 and Kodak Ektachrome 100S (the new
one). Than I mixed all the pictures before projecting them. I failed to
identify in a statistically valid manner the M or the R pictures.
Then I went to the testwall and shot a new series of pictures (again on
RSX50). Under the microscope (50xenlargement)I could see some differences
in flatness of field and micro contrast. The projection howver did fail to
show it.
Then I used the optical bench (from leica). Here I noticed a more
pronounced curvature of field and some slight astigamatism in the outer
zones. The resolution figures of the M version were also slightly better in
the corners and the outer zones.
The people of Leica provided me with the MTF graphs and again the
microcontrast of the M is better, translating in a better retention of
sharp details in the very fine image structures.
So I put some Fuji Velvia in the cameras, shot again a number of
transparancies, now under carefully controlled outdoor shooting (overcast
sky, heavy tripod, calibrated handheld meter etc). And now I was able to
see the difference: the M lens is indeed slighty sharper, but you need an
enlargement factor of more than 15x to see it. The R lens on the other side
while losing some microdetails had a smoother representation of the object
structures. However: 1/3 of a stop from the optimum exposure (for the
Velvia) ruined the micro contrast and the resolution of the fine details.
The lesson: it takes some work to push both lenses to the outer limits of
performance. And the differences while clearly visible are in a different
direction. The best lens in this case is not easy to identify. The ultimate
in sharpness is provided by the Summicron-M, the smoothest image by the R.
The ideal lens still has to be computed.
Maybe this story gives you some food for thought.
Erwin Puts